Andrew Kaster

Sounds like the problem with getaddrinfo is still there. Still not sure what we're doing wrong with those, though I haven't looked into it very much.

Re: librt, does this patch do the trick? ``` diff --git a/Meta/Lagom/CMakeLists.txt b/Meta/Lagom/CMakeLists.txt index dc89743d2a..d7a5023bed 100644 --- a/Meta/Lagom/CMakeLists.txt +++ b/Meta/Lagom/CMakeLists.txt @@ -360,7 +360,8 @@ add_serenity_subdirectory(Userland/Libraries/LibCore) target_link_libraries(LibCore PRIVATE Threads::Threads) if (${CMAKE_SYSTEM_NAME}...

Re: addrinfo: Does this patch fix it? ``` diff --git a/Userland/Libraries/LibCore/System.h b/Userland/Libraries/LibCore/System.h index a57d1f8faf..f190672f71 100644 --- a/Userland/Libraries/LibCore/System.h +++ b/Userland/Libraries/LibCore/System.h @@ -263,7 +263,7 @@ private: } struct AddrInfoDeleter { - void...

We bumped our minimum required gcc version to 13.x recently :sweat_smile: . That or clang-17+

Ah hmm. Does qualifying that call as AK::popcount fix the build error?

cc @clausecker

An LLVM PR was created that should resolve this on trunk. Not sure if it'll end up getting back ported to the 18 release branch -- we could always ask...

Hm. This appears to be in the SIMD Wasm implementation. I suspect that trying to load wasm programs using that extension to the wasm spec on a 32-bit platform without...

I'll merge this, but I am a bit nervous about early panics that are hard to debug.

Thanks for fixing this! A couple comments: - Can you add one of the reproducers from the issue as a test? A Layout test seems appropriate, given the last example...